Taking the plant community of Hefei Botanical Garden as the research object, semantic segmentation method is used to evaluate the scenic beauty of the plant community. It points out the importance of scenic beauty estimation of plant communities and the problems with existing methods. Subsequently, a review of the current research status both domestically and internationally is conducted,  and existing research methods and achievements are summarized. Next, it explains the content and technical route of this study, including data sources and processing, landscape image segmentation of plant community, color feature extraction, texture feature extraction, and establishment of scenic beauty estimation models. Finally, the effectiveness and feasibility of the proposed scenic beauty estimation model are verified through experiments. This paper has guiding significance for improving the landscape design and management level of botanical gardens, and can also provide reference and guidance for research in other related fields.